Controlling: the composition of plasma-activated water by Cu ions

The colloidal solution is synthesized by laser ablation of Cu target (purity: 99.9%) in purified water. The experimental setup has been described in details in [42]. Accordingly, the target is immersed in a beaker containing water and kept at constant depth (2 cm) during the synthesis, in order to keep the laser-ablation efficiency constant [43]. The 1064 nm Nd:YAG laser (Brilliant, Quantell) is focussed on the target, that is ablated with 10 000 pulses in 50 ml of water. The pulse duration is 4 ns with a repetition rate of 5 Hz. The energy delivered to the target surface is 120 mJ, while the beam width on the target surface is 338 μm, which yields a laser pulse fluence of 34 J cm⁻². From the volume of the craters left on the surface after ablation the amount of ablated material forming the colloidal solution can be determined [44, 45]. The volume of the ablated material is found to be 20 × 10⁶ μm³, which corresponds to 0.51 mg of ablated materials per crater (or per 50 ml solution), which yields the mass concentration of ablated species to be 10⁻² mg ml⁻¹.

The synthesized colloidal solutions are characterized by optical absorbance measured with the UV–Vis spectrophotometer (Lambda 25, Perkin Elmer) in the 190–800 nm wavelengths range. The morphology and the size-distribution of the CuNPs in the colloid are studied by scanning electron microscopy (SEM) images (Jeol JSM-7600F). Few drops of the CuNP-colloid are dropped onto a Si substrate, and analysed by SEM after the water evaporation. The size-distribution shows a log–normal distribution (figure 1(a)), which is typical for the NPs synthesized by laser ablation in water [45, 46], with the average CuNP diameter of 38 nm.

Zoom In Zoom Out Reset image size

The SEM image (figure 1(b)) shows that additionally to the CuNPs, a certain amount of ablated material formed irregular surfaces (flakes) rather than NPs. The crystallinity and phase of the CuNPs are determined using Siemens D5000 x-ray diffractometer with Cu Kalpha-radiation and Göbel mirror in grazing incidence (GI) geometry. The GIXRD measurements are performed with the fixed incidence angle of 1°. According to the GIXRD spectrum shown in figure 1(c), the diffraction peaks match to the Cu₂O phase, showing that oxidized rather than pure CuNPs are formed. The zeta potential (ζ) of the Cu₂ONPs is determined by electrophoretic light scattering, using the Zetasizer Ultra (Malvern Panalytical, UK). ζ is calculated from the measured electrophoretic mobility by means of the Henry equation using the Smoluchowski approximation (f(ka) = 1.5) with the ZS Xplorer 1.20 (Malvern Panalytical). The ζ is found to be 13.7 ± 0.6 mV. In accordance to this, the colloid has been stable for several weeks.

Active species are generated in purified water (produced with the ELGA Purelab Option-R 7 purifier), copper colloidal and salt solutions by putting the liquid surface in contact with the plasma plume of a kHz plasma jet and of a surface-wave microwave discharge, respectively.

The surface-wave microwave discharge is generated with the help of a surfatron launcher (Sairem, Surfatron 80) in a quartz tube of outer diameter 7 mm and inner diameter 4 mm, using Ar gas at 2000 sccm flow rate and MW input power of 25 W. A Berzelius beaker of 35 ml filled with 32 ml of liquid is positioned below the plasma plume with the liquid surface being at distance d (for each sample specified in section 3) from the edge of the quartz tube [41]. In order to avoid the overheating of the quartz tube during treatments, compressed air is applied along the quartz tube with a gas flow rate of 8 slm.

The atmospheric pressure plasma jet consists of a quartz tube with the outer and inner diameters of 1.5 and 1 mm, respectively, and a copper wire of 100 μm diameter, which is inserted in the capillary and serves as an electrode [47]. The electrode is powered with a sinusoidal voltage waveform of 28 kHz with 12 kV maximum voltage, chosen based on our previous optimization work conducted on the single electrode atmospheric pressure plasma jet [48]. The discharge is generated in N₂ (99.996% purity), which is supplied to the capillary with 500 sccm flow rate. The 25 ml sample in the Berzelius beaker is brought in contact with the plasma jet by placing the liquid surface at 5 mm distance from the capillary orifice, which insures the comparable production of H₂O₂ and ${\mathrm{N}\mathrm{O}}_{2}^{-}$ in the PAL [41].

The plasma-activated samples are analysed immediately after the treatment, as well as several times during their storage. The concentration of ${\mathrm{N}\mathrm{O}}_{2}^{-}$, ${\mathrm{N}\mathrm{O}}_{3}^{-}$ and H₂O₂, and the pH of samples are measured with QUANTOFIX^Ⓡ test strips (Nitrate/nitrite 500, 10–500 mg ${\mathrm{l}}^{-1}\enspace {\mathrm{N}\mathrm{O}}_{3}^{-}$, 1–80 mg ${\mathrm{l}}^{-1}\enspace {\mathrm{N}\mathrm{O}}_{2}^{-}$; Nitrate/nitrite 100, 5–100 mg ${\mathrm{l}}^{-1}\enspace {\mathrm{N}\mathrm{O}}_{3}^{-}$, 0.5–50 mg ${\mathrm{l}}^{-1}\enspace {\mathrm{N}\mathrm{O}}_{2}^{-}$; Peroxide 25, 0.5–25 mg l⁻¹; Peroxide 100, 0.5–25 mg l⁻¹). The strips are evaluated with the QUANTOFIX^Ⓡ Relax unit (by Macherey-Nagel, GmbH), which allows quantitative analysis with high accuracy. The Nitrate/nitrite strips have been calibrated using NaNO₂ and Ca(NO₃)₂ solutions of known concentrations, and the calibration has been verified with UV–VIS absorption spectroscopy measurements. Using strips makes possible to follow the ageing of samples with 1 min time resolution without any significant consumption of the sample (i.e. the sample volume can be kept quasi-constant during ageing). The measuring error has been determined to be less than 10% [41].

First of all, the effect of a bulk copper surface is validated by using a copper foil. During the treatment the inner wall of the Berzelius beaker is wrapped with copper foil (the size of the foil is 4.5 × 10 cm). After treatment this foil is transferred to the storing container together with the PAW. As a control sample, a plasma-activated water (PAW1) is also produced. The samples are prepared with the Ar surface-wave microwave (MW) discharge (see section 2.2) using treatment time of 5 min, and treatment distance of 8 mm (as listed in Table 1), which according to the previous studies [41] results in PAL with relatively high H₂O₂ concentration, such as 40–50 mg l⁻¹.

The first column of figure 2 shows the composition and the pH of the two samples, as well as their evolution during storage. After the treatment, in the case of the PAW1 the ${\mathrm{N}\mathrm{O}}_{2}^{-}$ concentration starts to decrease strongly, and as a consequence, the ${\mathrm{N}\mathrm{O}}_{2}^{-}$ disappears totally at about 200 min of storage. This occurs due to the high H₂O₂ concentration, since as already discussed in [41], the rate of the concentration decrease depends on the absolute concentrations, most importantly on that of the H₂O₂. Meanwhile, the ${\mathrm{N}\mathrm{O}}_{3}^{-}$ concentration slightly increases, since ${\mathrm{N}\mathrm{O}}_{3}^{-}$ is one of the possible end products of the ${\mathrm{N}\mathrm{O}}_{2}^{-}$ + H₂O₂ + H⁺ → products reaction. Due to this reaction, the H₂O₂ concentration also decreases during the first 100 min, after which stabilizes at about half of its initial value. When applying the copper foil (PA(W + Cu)1 sample), contrary to the PAW1 case, the H₂O₂ concentration does not stabilize after 100 min of storage, but decreases further with an even higher rate, and as a consequence, H₂O₂ disappears on the 2nd day (at 2000 min) of storage. On the other hand, the nitrite concentration stays quasi-stable up to one week (10⁴ min). The pH of the liquid shows a strong correlation with the H₂O₂ concentration, i.e. the pH increases monotonically with the decrease of the H₂O₂ concentration. Although the initial pH of the PAW1 and PA(W + Cu)1 are both close to 5.5, their ageing is different. While the pH of the PA(W + Cu)1 increases to 6 during one week of storage, the pH of the PAW1 decreases to 4.5.

Zoom In Zoom Out Reset image size

The effect of the copper foil is attributed to the copper ions participating in the Fenton-type reaction (similar effect is applied for OH production aimed for Tungsten removal in [52]), which is believed to have more pathways: (i) the Fenton Cu⁺/Cu²⁺ + H₂O₂ → Cu²⁺/Cu³⁺ + OH + OH⁻ or (ii) the catalase Cu²⁺ + H₂O₂ → CuO²⁺ + H₂O, CuO²⁺ + H₂O₂ → Cu²⁺ + H₂O + O₂. In the case of acidic conditions the OH⁻ produced in the Fenton reaction can react with the H⁺, i.e. OH⁻ + H⁺ → H₂O, which results in the increase of pH.

The Fenton-type reaction can occur on the surface of the copper foil or in the liquid phase due to the dissolved copper ions. Removing the copper foil the surface effect can be eliminated. By removing the copper foil (PA(W + Cu)1-Cu sample) before the second phase of H₂O₂ concentration decrease, the slowing of this phase is achieved, while the nitrite concentration decrease enhances, as shown in the first column of figure 2. As a consequence, after one week (10⁴ min) of storage both the nitrite and the H₂O₂ are still present in the plasma-activated liquid. By removing the copper foil, the dissolution of Cu-ion is also stopped. The concentration of dissolved Cu-ion after one week of storage in the PA(W + Cu)1 is measured to be above 30 mg l⁻¹, while in the PA(W + Cu)1-Cu it is around 30 mg l⁻¹. From the evolution of the species concentrations for the three cases presented in the first column of figure 2, it can be concluded that, with the presence of the Cu-ions in the system, there is a competition between the Fenton-type reaction and the ${\mathrm{N}\mathrm{O}}_{2}^{-}$ + H₂O₂ + H⁺ → products reaction controlled by the active species concentrations ⁵ .

The effect of the Cu ions originating from CuNPs is tested by using CuNP-colloidal solution in two different ways: (i) by treating directly the CuNP-colloid, i.e. creating PAC and (ii) by adding the colloidal solution (COL) to the PAW.

In the first case, the colloid is directly treated with the plasma plume under the same conditions as in the previous experiment. Comparing to the plasma-activated water, PAW1, the obtained PAC1 has a similar composition, however a slight enhancement in the species production can be observed. At the same time, during ageing, the species concentrations in the PAC exhibit different behaviour than in the PAW1, as shown in the second column of figure 2. In the case of the PAC1, the ${\mathrm{N}\mathrm{O}}_{2}^{-}$ concentration, after a slight decrease occurring during the first 10 min, remains stable for about 2 days (2000 min). Following this, the ${\mathrm{N}\mathrm{O}}_{2}^{-}$ concentration starts to decrease, and finally ${\mathrm{N}\mathrm{O}}_{2}^{-}$ disappears on the 7th day (10⁴ min) of storage. Meanwhile, the ${\mathrm{N}\mathrm{O}}_{3}^{-}$ concentration slightly increases and reaches the same value as in the case of the PAW1. The H₂O₂ concentration exhibits a similar behaviour as in the case of the PAW1, i.e. decreases during the first 100 min and quasi-stabilizes at about half of its initial value. The difference between the PAW1 and PAC1 lies also in the pH of the liquids, the PAW being more acidic, 5.5 versus 6.2. The pH of the PAW1 is quasi-stable up to 400 min, after which it monotonically decreases and reaches pH 4.5 at the 7th day (10⁴ min) of storage. In the case of the colloid, the acidification starts with the decrease of the nitrite concentration. After one week of storage the pH of the PAC1 reaches the initial pH of the PAW1.

Comparing the colloid and copper foil cases, although the copper ions that mediate the H₂O₂ reaction pathway are present in both liquids, the evolution of the species concentrations during storage is different. While in the case of the PA(W + Cu)1 the H₂O₂ concentration monotonically decreases to zero, in the PAC1 the H₂O₂ concentration stabilizes after 100 min, as in the case of the PAW1. As a consequence, the nitrite concentration in the PA(W + Cu)1 is quasi-stable during one week (being stable even up to 45 days at [${\mathrm{N}\mathrm{O}}_{2}^{-}$] = 5.6 mg l⁻¹) and the nitrate concentration slightly decreases to [${\mathrm{N}\mathrm{O}}_{3}^{-}$] = 30 mg l⁻¹. On the other hand, in the case of the PAC1 the nitrate and nitrite concentrations reach the same values as in the PAW1 with a one week delay. In the two cases the dissolved Cu-ion concentrations are found to be different. In the initial colloid the measured concentration is about 10 mg l⁻¹, which increases with plasma treatment and peaks at around 30 mg l⁻¹ after few hundred minutes, while in the case of copper foil the Cu-ion concentration peaks above 30 mg l⁻¹.

For the second colloid test case, the non-treated colloid is added to the PAW1. Two different mixtures of 32 ml are created, as follows: (i) the PAW1+COL with 16 ml PAW1 and 16 ml COL, and (ii) the PAW1 + W with 16 ml PAW1 and 16 ml W, which serves as a control sample. Due to the lower species concentrations obtained through dilution, and most importantly that of the H₂O₂, the ageing dynamics also changes [41], as shown in the third column of figure 2. In both mixtures the concentrations decrease at a lower rate comparing to the PAC1 and the PAW1, respectively. As a consequence in the PAW1 + W the ${\mathrm{N}\mathrm{O}}_{2}^{-}$ can survive 24 h longer than in the PAW1, while the H₂O₂ concentration is about half of the one obtained in the PAW1. In the PAW1 + COL sample the nitrite concentration decreases slightly during 6 days (8600 min) of storage and reaches a higher value than in the case of the PAC1, while the H₂O₂ and the nitrate concentrations are about half of the concentrations achieved in the PAC1. The more moderate concentration changes in the PAW1 + COL can also be attributed to the lower 10 mg l⁻¹ Cu-ion concentration.

Comparing to the PAW and the PAC, in the diluted samples no first phase concentration decrease is observed. This may be explained with the fact that by dilution the temperature of the sample decreases fast, while the undiluted samples cool down slower to room temperature from the 40^o C that are reached with the plasma treatment. As a consequence, in the case of the PAW1, PAC1 or PA(W + Cu)1, during the first 10–20 min the reaction rates are higher due to the elevated temperature.

When the role of the NPs in the plasma-activated liquid is considered to be only the control of the ageing process, and the liquid is meant to be used without NPs, it can be filtrated. The effect of filtration is tested on a plasma-activated colloid, PAC1s, produced under the same treatment conditions as PAC1. Additionally, during treatment magnetic stirring is also applied to the colloid, due to which the disturbance of the surface occurs and results in the increase of the treatment distance. As a consequence, the concentration of species in the PAC1s are lower than in the PAC1, thus the ageing dynamics is different (as previously shown in [41]).

Using a syringe filter housing 0.2μm pore membrane, 10 ml of colloid is filtrated at 1 min and at 10455 min, respectively. Figure 3 shows the evolution of the species concentrations and the pH of the original and the filtrated colloids.

Zoom In Zoom Out Reset image size

The original PAC1s is quasi-stable during the first week of storage, the concentrations decrease considerably only after 10⁴ min. Nevertheless, in the following 3 days (4000 min), the nitrite and H₂O₂ concentrations decrease only by 25%. The colloid filtrated at 10⁴ min follows the behaviour of the original colloid, except for the pH, which slightly decreases. On the other hand, the colloid filtrated at 1 min of storage is quasi-stable only up to 10³ min, and the ${\mathrm{N}\mathrm{O}}_{2}^{-}$ disappears at about 10⁴ min of storage. The Cu-ion concentration measured in the colloid filtrated at 1 min of storage is 10 mg l⁻¹, while in the PAC1s it is around 30 mg l⁻¹. Overall, the filtration does not influence the concentration of active species, the filtrated colloids follow the behaviour of the original colloid for a reasonable time, which allows their usage.

The effects achieved in the liquids treated with the surface-wave microwave discharge are also validated in the case of a nitrogen kHz plasma jet (section 2.2), which generally, comparing to the surface-wave microwave discharge, produces lower H₂O₂ concentrations and [H₂O₂]/[${\mathrm{N}\mathrm{O}}_{2}^{-}$] in the PALs [41], and also results in lower PAL temperature of 30^o C. Figure 4 shows the composition of the different PALs created with the plasma jet, and their evolution during storage. The molar concentration of the H₂O₂ and ${\mathrm{N}\mathrm{O}}_{2}^{-}$ in these PAL are quasi-equal, with the H₂O₂ concentration being slightly higher. As a consequence, due to the ${\mathrm{N}\mathrm{O}}_{2}^{-}$ + H₂O₂ + H⁺ → products reaction, the two concentrations change similarly over time during storage. This is well illustrated by the plasma-activated water (PAW3) shown in the first column of figure 4. Due to to the slightly higher initial H₂O₂ concentration, the ${\mathrm{N}\mathrm{O}}_{2}^{-}$ recombines totally, while the H₂O₂ is still present in the liquid after 10⁴ min of storage. Contrary to the PAW3, in the case of the plasma-activated colloid (PAC3) (shown in the second column of figure 4), after 200 min the Fenton-type reaction becomes dominant over the above mentioned reaction, which results in the ${\mathrm{N}\mathrm{O}}_{2}^{-}$ concentration to decrease at a lower rate. As a consequence, after 10⁴ min both the ${\mathrm{N}\mathrm{O}}_{2}^{-}$ and H₂O₂ are still present in the PAC3, and their concentrations are similar. Overall, the ${\mathrm{N}\mathrm{O}}_{3}^{-}$ concentration follows the same trend as in the case of the surface-wave microwave discharge activated water and colloid, respectively.

Zoom In Zoom Out Reset image size

When using copper foil instead of NPs, the concentration of species show very similar behaviour during ageing (first column of figure 4), however, in the case of the H₂O₂ concentration the decreasing rate is faster. Comparing to the MW sample, although the initial H₂O₂ concentration in the PA(W + Cu)3 is about 3 times lower than in the case of MW discharge activated sample, the H₂O₂ disappears from the liquid on the same timescale. Due to the lower concentrations, in order to preserve H₂O₂ in this case the copper foil should be removed earlier than in the case of MW sample. Concerning the removal of the NPs, the PAC3 filtrated at 10 min of storage shows very similar ageing dynamics as the non-filtrated PAC3. The efficiency of nanoparticle filtration is tested by treating a filtrated colloid. The species concentrations in the plasma-activated filtrated colloid (PA(filtratedC)) follow the same trend during storage as in the PAW3 (shown in figure 4), which indicates the efficiency of the filtration. This is also confirmed by the UV–VIS absorption spectra shown in figure 5, where the two spectra overlap each other.

Zoom In Zoom Out Reset image size

Figure 5 further shows the absorption spectra of the CuNP-colloid, the plasma-activated colloid (PAC3) and PAW3 produced with the N₂ plasma jet and measured after 10 min of storage. The UV–Vis spectrum of the CuNP-colloid shows transparency in the visible (not shown here), and high absorbance in the UV part of spectrum, exhibiting a weak shoulder at 330 nm (blue-shifted from visible, typical peak of pure Cu nanoparticles (NPs) occurs at around 600 nm [49]), which is a typical spectrum for the Cu-oxide NPs. It is known that the ablated Cu oxidizes due to the interaction with dissolved oxygen in water [50]. The absorption spectra of the filtrated colloid indicates a very efficient filtration, as no significant absorption is observed in the relevant spectral range. During plasma treatment one observes the colour change of the colloidal solution (from greenish to transparent), which indicates the agglomeration of NPs. The change of the colloid is also illustrated by the UV absorption spectra, showing the decrease of the absorbance above 225 nm. Meanwhile, due to the creation of different species by the plasma treatment, a strong absorption occurs in the 190–240 nm spectral range. The UV absorption in the plasma-activated waters has been analysed in details by Oh et al [51], and the absorption in the 190–240 nm spectral range has been attributed to the absorption of the dissolved O₂, and the created H₂O₂, ${\mathrm{N}\mathrm{O}}_{2}^{-}$ and ${\mathrm{N}\mathrm{O}}_{3}^{-}$ species. In the present samples there is no significant dissolved O₂ (since Ar and N2 discharges have been used), thus the absorption can be attributed to the created H₂O₂, ${\mathrm{N}\mathrm{O}}_{2}^{-}$ and ${\mathrm{N}\mathrm{O}}_{3}^{-}$ species.

The effect of the Cu ions originating from salts is tested by using 0.5 mM copper acetate (Cu(CH₃COO)₂) and copper chloride (CuCl₂) solutions. The composition of the salt solutions treated with the Ar surface-wave microwave discharge at the same conditions as the colloidal solutions (as listed in table 2) are shown in figure 6, in comparison with that of the plasma-activated CuNP-colloids and PAW, respectively. The first salt sample is a plasma-activated copper acetate solution, PA(CuAc)1, created under the same conditions as the PAC1. One obvious difference between the two samples, as shown in the first column of figure 6, is the higher species concentrations obtained in the PAC. The salt solutions are prepared with a copper ion concentration of 32 mg l⁻¹, which in the PAC is measured after 200 min ageing, while in the initial CuNP-colloid the Cu-ion concentration is about 10 mg l⁻¹. The higher Cu-ion concentration during plasma treatment can result in the lower H₂O₂ concentration in the treated salt solution.

Zoom In Zoom Out Reset image size

The PA(CuAc)1 also exhibits a slightly different ageing dynamics. The strong first phase concentration decrease characteristic for the PACs and the PAWs during the first 100–200 min of storage, in the case of the PA(CuAc)1 is more moderate and it occurs only in the case of H₂O₂. This can be attributed to the higher Cu-ion concentration in the PA(CuAc)1 in this phase, due to which, the reaction kinetics is dominated by the Fenton-type of reaction over the reaction of H₂O₂ with nitrite ions. After the first phase, the two liquids show similar behaviour over time, and at about 5 × 10³ min of storage the nitrite and H₂O₂ concentrations, respectively, become quasi-equal. Meanwhile, the nitrate concentration is about 30% higher in the PAC1. At this point the main difference between the two liquids is the pH, which is about 4.5 in the PA(CuAc)1 and 5.5 in the PAC1. The effect has also been validated on liquids treated at a higher 9 mm distance, the PA(CuAc)2 and the PAC2, as shown in the second column of figure 6 (illustrating the reproducibility of the method).

Due to the degradation products of salts and their reaction with the active species of interest, a careful choice of the salt type is needed. While with the application of the CuAc similar effects can be achieved as with the CuNP-colloid, by choosing CuCl₂ the concentration of active species behave differently (third column of figure 6). Although the ageing of the PA(CuCl₂) is similar to that of the PAW1, the reaction kinetics is more complex, and additionally under acidic conditions the formation of hydrogen chloride can also be expected. By treating copper acetate solution with the plasma jet similar ageing behaviour is obtained as in the case the MW treated solution. The evolution of the species concentrations over storage time is shown in figure 7 in comparison with the case of the treated colloid.

Zoom In Zoom Out Reset image size

Overall, the CuNP-colloid and CuAc have similar effect on the PAW composition. However, while the CuNPs can be filtrated from the colloid, as proven by the UV–VIS absorption spectra shown in Figure 5, in the case of the PA(CuAc) different compounds created through plasma decomposition of acetate are also present. Comparing the absorption spectra of the PA(CuAc) to that of the PAW, significant differences are found at spectral ranges below 210 nm and above 230 nm, see figure 8. The copper acetate solution shows a strong absorption below 240 nm. However, this contribution is much lower in the treated solution, which indicates that the acetate is decomposed during plasma treatment.

Zoom In Zoom Out Reset image size